marcwilson wrote:
just a thought..if schenider can make their 28mm Pc in differnent mounts.canon nikon, etc why could zeiss not have taken this approach?
Maybe Zeiss didn't *want* to have their lenses scrutinized on a 5D sensor.
Canon has micro motors to move focus and aperture blades, that s a whole lot of more technology and as we see the Contax N lenses did change design to be able to use this technology
I agree with LeSong. I struggled to see the areas where the Zeiss won, of course we are looking at web jpgs. Maybe the raws would make it clearer. However, the areas where Canon won was obvious even looking at web jpgs.
I doubt the lens was designed for a Nikon cropped digital SLR. Zeiss public position is digital isn't there yet. Also, Zeiss talks about Canon customers in a Q&A and points to the ZS.
The areas where the adaptor might have influenced the results were the close focusing and CA. I don't think we have ever seen corner resolution influenced by an adaptor. Even if the CA and close focusing was resolved, I doubt it would tip the scales, would it? Hopefully someone can do some FF film test and let us know.
That leaves either sample variation and/or Cosina's skills.
Does this drive the price of the CZ21 up ?
Edited by jonboring on Jan 29, 2007 at 12:54 PM GMT
jonboring wrote:
I agree with LeSong. I struggled to see the areas where the Zeiss won, of course we are looking at web jpgs. Maybe the raws would make it clearer. However, the areas where Canon won was obvious even looking at web jpgs.
I doubt the lens was designed for a Nikon cropped digital SLR. Zeiss public position is digital isn't there yet. Also, Zeiss talks about Canon customers in a Q&A and points to the ZM.
Regardless of whether it is better or worse than the Canon 24mm, there is the more serious question of why the observed performance of the ZF 25mm in this test so significantly departs from Zeiss' reliable, published MTFs.
jonboring wrote:
The areas where the adaptor might have influenced the results were the close focusing and CA...
These are exactly the areas where the thickness of the adapter will absolutely has no influence because the ZF 25 has a linear focus group. If the adapter is too thick, one looses infinity. If it is too thin, it will loose a little bit close focus ability. But in both case, the image quality is abosolutely not affected by the thickness of the adapter alone.
On the other hand, if the thickness of the adapter ring is not even, then a tilted adapter could cause blurred corners, but it won't introduce CA. Because once the lens is tilted, the whole group is tilted together. The strong Longitudinal CA of the ZF25 as shown in Mark's test at large apertures could only be caused by individual elements that are not properly aligned with the rest of the group.
Mark, thank you for another excellent and highly interesting test. It is really surprising that this new, slow, big and impressive-looking design by Zeiss didn't perform better. I agree with LeSong that it does not look like an adapter issue, and maybe we're looking at a QC issue in which case it is even worse. Or are we having too high expectations?
I'm ordering the Canon 24L now, when it arrives I'll make a careful test to see if I can duplicate your findings as to its performance characteristics. Should be fun.
My first instincts were that the adaptor might be at fault, but the more time I spent with the lens, the less the adaptor seemed to blame. The same switch from thick to thin adaptor transforms the performance of the Nikon 15mm and 17-35mm, but it makes no difference to the 28mm AIS or this ZF25mm. I'm awaiting Zeiss' response before coming to any hard and fast conclusions but at this point it seems unlikely to me that the adaptor is the problem.
Where the resolution of both lenses appears to be very close, as it was throughout the centre frame samples, it still seemed to me that the ZF consistently produced more attractive images. I think the final score 13/12 reflects that fairly.
It's a wiki world: please everybody test your ZF25s and ZF85s and N17-35s and let's see what happens. Even if my testing procedure is utterly flawless, it only evaluates one sample. When that one sample appears to fall below expectations, the onus falls on everyone to better it. It would only take one test in which the ZF25 convincingly outperforms a healthy 24L to prove that the design is capable of more. As I've always said: a bad result can have any number of explanations, but a good one is irrefutable.
If this does turn out to be a Cosina QC problem (and that seems where it's headed) then it doesn't bode well for the rest of the line, as one of the main marketing points is the high quality construction.
In fact I believe they have delayed production of the ZM 85mm f2 Sonnar for a major redesign, because the internal focussing system was deemed inadequate for long term durability.
Zeiss are usually pretty tough on QC, lets hope they come back with a good answer.
I think we should all thank Husband and test our own. If they indeed all look like this then Zeiss has falsly advertised the MTF and I want mine fixed.
I am testing against the canon zooms becuase for me a result aginst a prime is not enough you must do far far better than a zoom to justify the extra weight.
For me my 25/28/35,35-70 only get used at F8 or f11 at worst and I expected better at f8 in the corner, this looks the same as my 30 year also 28mm f2.8.
But in the center using a Nikon as I have tested you sure need a better sensor
Although this doesn't affect me personally, being a Leica owner, I can't help but be very disappointed by these results. It would appear that we have partly lost one of the great brands in photography.
On the other hand, as some others have observed, it appears that Canon may finally have gotten a grip on their QA and are putting out pretty solid lenses, even wide-angles ones. The 24L probably is not as good as the 35L, but I think it seems good enough to extend Canon's range of acceptably good lenses by quite a bit.
Actually as a Leica user you are affected by Zeiss lenses, they make some decent glass for the M mount, read Puts.
Canon makes decent glass, I miss L glass quality in their less fast lenses and the 50 and UWA are not covered with the same high qualtiy lenses like the 1.4/35/24/. I loved the c/y lenses for their quality and their (used) price, if I have to put the same amount of money like for a Canon L superfast lens and loose aperture control, I want it to be top notch.
I think LotusM50 has rised a good question in comparing the results with the MTF and I wonder if this is indeed a Cosina QC problem.
Do the Cosina-Zeiss ZM lenses suffer from the same lack of accordance with the specs or do they have any other known problems?
Bruce Connelly just mailed me to say that his ZF25 appears to be performing much better than mine. I'm waiting for pix; no doubt he'll chime in shortly. It would be very disheartening to see such a great brand blighted by Cosina-style sample variation.
With Zeiss and Leica you sense that every item emerging from the factory gates is as the designer intended.
carstenw wrote:
Although this doesn't affect me personally, being a Leica owner, I can't help but be very disappointed by these results. It would appear that we have partly lost one of the great brands in photography.
I think that conclusion is quite premature.
I will try to get the 2 Japanese magazine reviews (one page, "first look" reviews from Asahi Camera and Nippon Camera) of the ZF 25mm that I have translated -- I posted a few statements from them about a months ago (which were quite positive) and sent scans of the reviews to someone who promised to translate and post the translation -- but that never materialized.
Lotusm50 wrote:
I think that conclusion is quite premature.
I will try to get the 2 Japanese magazine reviews (one page, "first look" reviews from Asahi Camera and Nippon Camera) of the ZF 25mm that I have translated -- I posted a few statements from them about a months ago (which were quite positive) and sent scans of the reviews to someone who promised to translate and post the translation -- but that never materialized.
Perhaps Carsten meant his statement in regards to Quality Control/Sample variation. If other samples do indeed show much better performance, then one can no longer depend on the name Zeiss. One thing which does surprise me if this does turn out to be a QC/sample issue is that just about every Zeiss M lens tested on the M8 has performed extremely well, often the equel of Leica in regards to resolution. In some instances, the Zeiss versions have actually outperformed Leica. At least that's what Sean Reids reviews seem to have shown. Perhaps these are simpler designs relying less on the SLR inverted telephoto issue for Wides(some are true wide angles) but QC would be extremely important as well. We do know that Zeiss actually builds the most complex wide angles with floating elements in Germany still(and they are priced accordingly) which does give credence to Cosinas inability to actually produce these more complex designs.
Lotusm50 wrote:
I think that conclusion is quite premature.
I will try to get the 2 Japanese magazine reviews (one page, "first look" reviews from Asahi Camera and Nippon Camera) of the ZF 25mm that I have translated -- I posted a few statements from them about a months ago (which were quite positive) and sent scans of the reviews to someone who promised to translate and post the translation -- but that never materialized.
My wife is Japanese and we are both avid photographers (semi-pro), so send them my way if you wish and we'll do our best.
According to Zeiss' PR person, one of the company's engineers is looking at the test today, so we should have 'official' feedback soon. The review could be repeated with a different sample, and/or a ZS vesion.
No doubt we'll see a clearer picture emerge as more lenses get into circulation. I'd like to think that a ZF25 would give better results that I was able to extract from this one.
If anyone out there uses a horsmean digiflex husband could test his zf lens on full frame..but I know only of pepole selling them on AOP site..not using any.
I do not think the zs version exists yet except for the 50mm lens.
I would love to say I'll pick up a 25mm tomorrow and test it on the 1ds but I have to wait a bit to get my wide prime.
On the Zeiss ZF25mm web page there is no mention of a floating element, whereas it's feature quite prominently in the copy for the ZF50/2 Makro Planar. The sample I tested, on both adaptors tried, suffered a resolution collapse in the corners at close range that would be consistent with the omission of a floating element. However, I've asked Zeiss the same question, and am awaiting a reply.
